My research on pore-scale immiscible displacement uses 4D x-ray microtomography combined with in situ flow to look into the evolving fluid distributions during multiphase immiscible fluid displacement.

Here we investigate three fundamental questions:


Question #1: How do dynamic flow processes such as snap off and piston-like displacement behave during scCO2-brine drainage and imbibition in the porespace of real rocks?

The conceptual model of roof snap off. Figure credit: Dr Matthew G. Andrew

Significance: During drainage in 2-D micromodels we have observed piston-like displacement, layer flow, snap-off, and reconnection events. However, this study is the first to observe pore-scale displacement events of supercritical CO2 in a real rock at reservior conditions during both drainage and imbibtion. Furthermore, we were able to observe and measure local curvatures in situ both before and after events on a pore-by-pore basis to understand the mechanisms underlying post-injection fluid configurations. This is important for predicting fluid movement during CO2 sequestration and enhanced oil recovery.

Results from Drainage: Quasi-static capillary pressure changes associated with snap-off and displacement events were found and measured. We provide the first evidence for non-local interface recession associated with a change in capillary pressure in a real rock at reservoir conditions. Furthermore, snap-off was observed, both in the throat through which non-wetting phase invasion occurred (Roof snap-off) and in throats far away from it (distal snap-off). It was found that distal (non-local) snap-off produced more persistent fluid configurations than Roof snap-off. Equilibrium capillary pressure changes associated with drainage events are not sufficient to explain the snap-off associated with these events, as the disconnected ganglia formed have lower capillary pressures than the connected non-wetting phase either before or after the jump. The snap-off may instead be due to significant dynamic pressure gradients generated during these events as fluid pressures and interface curvatures attempt to re-equilibrate after rapid fluid movement. Snap-off in these throats is not only controlled by throat radius and proximity to the event but also by the local fluid arrangement. A more mobile local wetting phase may lead to a greater likelihood for a throat to snap-off during an event.

a-d Curvature maps for the connected and disconnected CO2 regions, rendered in a subvolume from four sequential images during a Roof snap-off sequence. Although a subvolume is used in a–d for clarity, curvature measurements for the connected CO2 were made across the entire CO2 surface. e The resulting capillary pressure distributions. Low capillary pressures in the disconnected phase region may be result of low transient capillary pressures generated during the drainage event. Figure Credit: Dr Matthew G. Andrew

Title: The Imaging of Dynamic Multiphase Fluid Flow Using Synchrotron-Based X-ray Microtomography at Reservoir Conditions

Authors: M.G. Andrew, H.P. Menke, M.J. Blunt, B. Bijeljic

Abstract: Fast synchrotron-based X-ray microtomography was used to image the injection of super-critical CO2 under subsurface conditions into a brine-saturated carbonate sample at the pore-scale with a voxel size of 3.64 μm and a temporal resolution of 45 s. Capillary pressure was measured from the images by finding the curvature of terminal menisci of both connected and disconnected CO2 clusters. We provide an analysis of three individual dynamic drainage events at elevated temperatures and pressures on the tens of seconds timescale, showing non- local interface recession due to capillary pressure change, and both local and distal (non-local) snap-off. The measured capillary pressure change is not sufficient to explain snap-off in this system, as the disconnected CO2 has a much lower capillary pressure than the connected CO2 both before and after the event. Disconnected regions instead preserve extremely low dynamic capillary pressures generated during the event. Snap-off due to these dynamic effects is not only controlled by the pore topography and throat radius, but also by the local fluid arrangement. Whereas disconnected fluid configurations produced by local snap-off were rapidly reconnected with the connected CO2 region, distal snap-off produced much more long-lasting fluid configurations, showing that dynamic forces can have a persistent impact on the pattern and sequence of drainage events.


Results from Imbibition: Imbibition appeared isotropic with no clear macroscopic gradients in saturation developed. Snap-off in this system appears to be an equilibrium process, with no observable difference in interface curvature between the connected and disconnected non-wetting phase regions generated after snap-off.

A-B: Connected and disconnected non-wetting phase (CO2) after invasion and snap-off during imbibition, taken from two subsequent tomographies. In this rendering the wetting phase (brine) and the rock grains are transparent. The surfaces are coloured according to the magnitude of the mean interface curvature. C: The distribution of curvatures, as measured on terminal menisci from the wetting-non-wetting fluid-fluid interface, identified using curvature anisotropy. The disconnected non-wetting phase region formed by snap-off has an indistinguishable interface curvature to the connected non-wetting phase region either before or after the snap-off event. Figure Credit: Dr. Matthew G. Andrew

Title: Dynamic Drainage and Imbibition Imaged Using Fast X-Ray Microtomography

Authors: M.G. Andrew, H.P. Menke, M.J. Blunt, B. Bijeljic

Abstract: Recent developments in X-ray microtomography have allowed for multiple fluids to be imaged directly in the pore space of natural rock with temporal resolutions of tens of seconds, allowing for the direct observation of pore-scale displacements . This technique was used to image, with a spatial resolution of 3.64 μm and a temporal resolution of 45 seconds, the injection of supercritical CO2 (drainage) and brine (imbibition) into a carbonate sample at conditions of pressure, temperature and salinity representative of subsurface flow (10 MPa, 50oC, 1.5 M) using a novel technique for flow control at extremely low flow rates. Capillary pressure was measured from the images by finding the curvature of terminal menisci of both connected and disconnected CO2 clusters, identified using local curvature anisotropy.
Snap-off in this system was examined by analysing a single snap-off event during drainage and imbibition. Capillary equilibrium concepts do not explain the low capillary pressures seen in the snapped off regions of the pore-space during drainage. The disconnected region created during drainage instead preserves the extremely low dynamic capillary pressures generated during a drainage event (Haines jump). Imbibition appeared isotropic with no clear macroscopic gradients in saturation developed. Snap-off in this system appears to be an equilibrium process, with no observable difference in interface curvature between the connected and disconnected non-wetting phase regions generated after snap-off.

Question #2: How does carbonate wettability impact the amount and distribution of CO2 after drainage and imbibition?

Initial and remaining CO2 and N2 saturations measured in this study, colored points, compared to saturations measured on the same samples at the core scale from *Al-Menhali and Krevor [2016].4 The solid lines show the best fit Land trapping model42 with their respective values for the parametrization constant, C. Figure Credit: Dr. Al-Menhali et al. 2016 Environ. Sci. Technol..

Significance: In this work we have observed the pore scale fluid displacement processes underlying the trapping of CO2 in carbonate rocks in the mixed-wet state characteristic of many oil fields. The work has direct implications for CO2 capture and storage in oil fields where capillary trapping is seen as a key mechanism underpinning the permanence of sequestered CO2. In a related work, observations reported in Al-Menhali and Krevor [2016] found that there was a significant decrease in trapping when measured over centimeter lengths of rock, amounting to the aggregate response of fluids in tens of thousands of pores. There was a much larger decrease in CO2 trapping than the N2 trapping. A hypothesis was proposed for the varying response of the fluids, relating changes in pore scale contact angle to the ratio of interfacial tensions in the fluid systems. In this work pore scale observations of the fluids were made to evaluate this hypothesis and further elucidate the processes underlying the decreased trapping.

Six contact angles measured in the CO2 phase (yellow lines) with their corresponding values in the brine phase (white text), showing a wide distribution of contact angle values for trapped CO2 in the mixed-wet limestone sample with (1 and 2) strongly water-wet, (3 and 4) intermediate-wet, and (5 and 6) wetting to CO2. Figure credit: Dr. Ali Al-Menhali

Title: Pore Scale Observations of Trapped CO2 in Mixed-Wet Carbonate Rock: Applications to Storage in Oil Fields

Authors: A.S. Al-Menhali, H.P. Menke, M.J. Blunt, and S.C. Krevor

Abstract: Recent developments in X-ray microtomography have allowed for multiple fluids to be imaged directly in the pore space of natural rock with temporal resolutions of tens of seconds, allowing for the direct observation of pore-scale displacements . This technique was used to image, with a spatial resolution of 3.64 μm and a temporal resolution of 45 seconds, the injection of supercritical CO2 (drainage) and brine (imbibition) into a carbonate sample at conditions of pressure, temperature and salinity representative of subsurface flow (10 MPa, 50oC, 1.5 M) using a novel technique for flow control at extremely low flow rates. Capillary pressure was measured from the images by finding the curvature of terminal menisci of both connected and disconnected CO2 clusters, identified using local curvature anisotropy.
Snap-off in this system was examined by analysing a single snap-off event during drainage and imbibition. Capillary equilibrium concepts do not explain the low capillary pressures seen in the snapped off regions of the pore-space during drainage. The disconnected region created during drainage instead preserves the extremely low dynamic capillary pressures generated during a drainage event (Haines jump). Imbibition appeared isotropic with no clear macroscopic gradients in saturation developed. Snap-off in this system appears to be an equilibrium process, with no observable difference in interface curvature between the connected and disconnected non-wetting phase regions generated after snap-off.


Question #3: How does oil rearrange in the pore space during drainage and imbibition?

Snap-off during imbibition at a pore junction – Case-I. Various time steps during brine injection showing displacement of oil and trapping: (a) t = 113 min 22 s (injected volume = 5.073 μL), (b) t = 121 min 36 s (injected volume = 5.442 μL), (c) t = 124 min 8 s (injected volume = 5.555 μL), and (d) t = 126 min 2 s (injected volume = 5.640 μL). Time ‘t = 0’ denotes the start of the imbibition process (after acquiring the first tomographic image in imbibition). Each time (t) represents the end of each tomographic image acquisition. (e) Location of the pore junction where the snap-off occurred. (f) Capillary pressure in the ganglion and connected side is plotted against time and injected volume. Figure Credit: Dr. Kamal Singh.

Significance: This study provides a pore-scale understanding of the dynamics of two-phase fluid flow during brine injection (imbibition) in a water-wet carbonate rock. We have used fast synchrotron X-ray micro-tomography to investi- gate fluid saturation, pore-scale brine-oil curvature, and pore-filling and snap-off events.
By analyzing the local capillary pressure calculated from brine-oil curvature data, we have investigated snap-off events with different local pore-space geometry and fluid configurations. The local capillary pressure on the ganglion side, where the oil becomes trapped, decreases at the onset of a snap-off event and creates a capillary pressure – and hence fluid pressure – gradient across the pore, driving the swelling of wetting layers until the threshold capillary pressure is reached and snap-off occurs. This process takes an order of 10 minutes, as opposed to the sub-second movement of fluid interfaces through the centers of the pore space previously observed in drainage. When a threshold pressure is reached, the brine-oil interface becomes unstable resulting in snap-off and trapping of the oil phase. After the snap-off event in imbibition, the oil re-arranges in the pore space to find a new position of minimum energy and the local capillary pressure rises, becoming higher than that in the connected non-wetting phase. On the other hand, a ganglion formed by Roof snap-off during drainage retains the capillary pressure that is lower than that of the local threshold capillary pressure just before the snap-off event: the reason again is that the trapped phase re-arranges itself in the pore space to minimize the oil pressure. We have also iden- tified pore-filling events during imbibition in which the local capillary pressure remains approximately constant while locally the oil saturation decreases, representing the rapid filling of a pore space once the wetting phase has traversed the threshold capillary pressure for filling.

Local capillary pressure and oil saturation analysis during pore-filling and snap-off processes. (a) Capillary pressure, calculated from the total curvature using the Young-Laplace equation, is plotted against time and injected volume in the ganglion and connected side subsets that are shown in Fig. 1. (b) Oil volume (v) normalized to the oil volume (v*) at t = 24 min 4 s in the ganglion-side subset and in a subset near throat (marked by a red box in the inset picture) as a function of time and injected volume. (c) Capillary pressure of the brine-oil interfaces in the ganglion-side subset as a function of normalized oil volume (v/v*) in the ganglion-side subset. Here, the error bars in the capillary pressure are standard error in the mean with 95% confidence intervals calculated using ±1.96 × σ/√N, where σ is the standard deviation and N is number of points sampled. Figure credit: Dr. Kamal Singh

Title: Dynamics of snap-off and pore-filling events during two-phase fluid flow in permeable media

Authors: Kamaljit Singh, Hannah Menke, Matthew Andrew, Qingyang Lin, Christoph Rau, Martin J. Blunt, Branko Bijeljic

Abstract: Understanding the pore-scale dynamics of two-phase fluid flow in permeable media is important in many processes such as water infiltration in soils, oil recovery, and geo-sequestration of CO2. The two most important processes that compete during the displacement of a non-wetting fluid by a wetting fluid are pore-filling or piston-like displacement and snap-off; this latter process can lead to trapping of the non-wetting phase. We present a three-dimensional dynamic visualization study using fast synchrotron X-ray micro-tomography to provide new insights into these processes by conducting a time-resolved pore-by-pore analysis of the local curvature and capillary pressure. We show that the time-scales of interface movement and brine layer swelling leading to snap-off are several minutes, orders of magnitude slower than observed for Haines jumps in drainage. The local capillary pressure increases rapidly after snap-off as the trapped phase finds a position that is a new local energy minimum. However, the pressure change is less dramatic than that observed during drainage. We also show that the brine-oil interface jumps from pore-to-pore during imbibition at an approximately constant local capillary pressure, with an event size of the order of an average pore size, again much smaller than the large bursts seen during drainage.